Distilling Out-of-Distribution Robustness from Vision-Language Foundation Models

TL;DR

This paper introduces a lightweight framework combining knowledge distillation and data augmentation to significantly enhance out-of-distribution robustness in vision models, leveraging robust foundation models as teachers.

Contribution

It demonstrates that large pretrained models serve as effective teachers for robustness and proposes Discrete Adversarial Distillation (DAD) using VQGAN for improved data augmentation.

Findings

Strong out-of-distribution robustness gains

Improved clean accuracy across architectures

Minor computational overhead

Abstract

We propose a conceptually simple and lightweight framework for improving the robustness of vision models through the combination of knowledge distillation and data augmentation. We address the conjecture that larger models do not make for better teachers by showing strong gains in out-of-distribution robustness when distilling from pretrained foundation models. Following this finding, we propose Discrete Adversarial Distillation (DAD), which leverages a robust teacher to generate adversarial examples and a VQGAN to discretize them, creating more informative samples than standard data augmentation techniques. We provide a theoretical framework for the use of a robust teacher in the knowledge distillation with data augmentation setting and demonstrate strong gains in out-of-distribution robustness and clean accuracy across different student architectures. Notably, our method adds minor computational overhead compared to similar techniques and can be easily combined with other data augmentations for further improvements.